N. Rajesh, M.R. Khare, and S.K. Pabi
Blast furnace, ANN modelling, silicon prediction, burden distribution, heat levels, scaffolding
Prediction of process parameters in a blast furnace (BF), which is a highly nonlinear, complex and continuous system, is a difficult task. Static models, which are based on first principles and tried in the past, are unable to give satisfactory results because of the dynamic nature of the process. In this paper artificial neural network (ANN) modelling of BF is explored. The concepts of ANNs are briefly discussed, and a few models reported in the literature are reviewed. The areas covered in this review are: (a) process control, viz. prediction of hot metal silicon, (b) process monitoring, viz. forecasting scaffold and scaffolding prediction of burden distribution, forecasting heat levels, (c) fault detection and diagnosis, viz. phenomena of the BF.
Important Links:
Go Back
